The customary method for the preparation of linear phosphonitrilic chloride polymers involves ring-opening polymerization of phosphonitrilic chloride trimer. Although workable, this method suffers from the fact that for satisfactory results to be achieved, highly pure cyclic phosphonitrilic chloride trimer must be used as the monomer. Such material is difficult and expensive to prepare.
Heretofore some work has been devoted to forming phosphonitrilic chloride polymers from lower molecular weight phosphonitrilic chloride oligomers. For example in J. Chem. Soc. 1960, 2542-7, Lund et al report an experiment in which a linear phosphonitrilic chloride oligomer of the formula (PNCl.sub.2).sub.11 PCl.sub.4.2 was heated with ammonium chloride in sym-tetrachloroethane under reflux. Polymerization occurred after 5.5 hours, at which time the amount of hydrogen chloride evolved corresponded to the composition (PNCl.sub.2).sub.10.6 PCl.sub.5. The rubbery product was extracted with light petroleum giving a significant quantity of a dark oil containing 10.5 percent PNCl.sub.2 trimer, the remainder of the oil consisting of cyclic polymers higher than the heptamer.
Moran in J. Inorg. Nucl. Chem. 30. 1405-13 (1968) investigated the thermal polymerization of the linear compound [Cl--(PCl.sub.2 .dbd.N).sub.3 -PCl.sub.3 ].sup.+ PCl.sub.6.sup.- in evacuated sealed tubes at 300.degree. C. for 5 hours and at 350.degree. C. for 5 hours. The phosphorus NMR spectrum of both samples indicated that polymers of other chain lengths were formed. The results in the 300.degree. C. case suggested to Moran that polymerization to the longer chain length compound [Cl--(PCl.sub.2 .dbd.N).sub.6 --PCl.sub.3 ].sup.+ PCl.sub.6.sup.- probably occurred. The NMR spectrum of the sample heated at 350.degree. C. indicated to Moran that polymers of both longer and shorter chain lengths were formed.
G. Allen et al in Polymer 11, 31-43 (1970) report attempts to prepare linear PNCl.sub.2 polymer by reacting PCl.sub.5 with ammonium chloride in ortho-dichlorobenzene, the ammonium chloride being introduced by stepwise addition to the reaction mixture. They were in hopes that the following reactions would occur: EQU (a) PCl.sub.5 +NH.sub.4 Cl.fwdarw.(1/n) Cl--(PCl.sub.2 .dbd.N).sub.n --PCl.sub.4 +4HCl EQU (b) Cl--(PCl.sub.2 .dbd.N).sub.n --PCl.sub.4 +NH.sub.4 Cl.fwdarw. EQU Cl--(PCl.sub.2 .dbd.N).sub.n --PCl.sub.2 .dbd.NH+3HCl EQU (c) Cl--(PCl.sub.2 .dbd.N).sub.n --PCl.sub.2 .dbd.NH+Cl--(PCl.sub.2 .dbd.N).sub.n --PCl.sub.4 .fwdarw. EQU Cl--(PCl.sub.2 .dbd.N).sub.2n --PCl.sub.4 +HCl
However they obtained very low molecular weight polymer (intrinsic viscosity of trifluoroethoxy derivative was below 0.05 dL/g). When they tried to increase the molecular weight of the polymer product by reacting it with NH.sub.4 Cl in o-dichlorobenzene solvent, they obtained a crosslinked material.
U.S. Pat. No. 3,443,913 discloses a method wherein linear (PNCl.sub.2).sub.3-15 oligomers are heated at 240.degree.-260.degree. C. to produce linear phosphonitrilic chloride polymers having a molecular weight between 3,000 and 10,000. However, this process involves heating for long periods of time, the endpoint of the polymerization occurring about 40 to 60 hours after heating has been initiated. The product obtained via this process is reported to be a dark orange viscous oil. See also James M. Maselli, Thomas Bieniek and Rip G. Rice (W. R. Grace and Company), Phosphonitrilic Laminating Resins, Air Force Materials Laboratory, Technical Report AFML-65-314; Wright-Patterson Air Force Base, Ohio: June, 1965, pages 18-19, which describes this same process. At page 47 of this report Maselli et al describe an experiment wherein oligomeric phosphonitrilic chloride was placed in a resin kettle fitted with a nitrogen inlet, stirrer and exhaust tube condenser. The kettle was heated to 250.degree..+-.10.degree. C. for a total of 55 hours while the polymeric (PNCl.sub.2).sub.n was stirred under a blanket of dry nitrogen. Samples of the reaction material were taken at selected intervals of time during the heating for molecular weight determination. The resulting data were as follows:
______________________________________ Time (Hours) Molecular Weight (VPO) ______________________________________ Start 700 10 1200 40 3200 55 6900 ______________________________________
According to the authors, when heating was continued for an additional 8 hours at temperatures in excess of 250.degree. C., the viscous, soluble oil (molecular weight 6900) was converted to the familiar insoluble "inorganic rubber".
In U.S. Pat. No. 3,545,942 which in part discloses a method of thermally stabilizing phosphonitrilic chloride oligomers by heating them in an inert atmosphere for 2 to 8 hours at 240.degree. to 260.degree. C., Rip G. Rice et al indicate that prolonged heating of the oligomer can result in the formation of an "inorganic rubber". A decade earlier Lund et al (op. cit.) referred to an experiment in which heating of a linear phosphonitrilic chloride oligomer in tetrachloroethane solution resulted in polymerization after 29 hours.
In prior applications Ser. No. 956,227 filed Oct. 30, 1978 and Ser. No. 176,926 filed Aug. 11, 1980, a distinctly superior thermal polymerization process is described wherein linear phosphonitrilic chloride oligomer is heated to 275.degree. to 350.degree. C. for 1 to 20 hours while concurrently withdrawing phosphorus pentachloride vapor from the liquid phase. A similar procedure is described in Japanese Laid-Open Application (Kokai) No. 55-27,344 published Feb. 27, 1980. In this case a linear phosphazene oligomer usually having a degree of polymerization of 3 to 15 is heated under reduced pressure (usually less than 20 mm Hg) to produce linear polymers. Heating for five hours or more at 100.degree.-300.degree. C. is suggested. Unfortunately, phosphorus pentachloride vapor is extremely corrosive at elevated temperatures--it tends to rapidly corrode even the most expensive corrosion-resistant metals used in the manufacture of corrosion-resistant chemical reactors.
Japanese Kokai No. 55-56,130 published Apr. 24, 1980 describes a method for producing phosphazene polymers in which a linear phosphazene oligomer is heated in the presence or absence of a solvent at 50.degree. to 300.degree. C. using a Lewis base such as urea, thiourea, polyurea or polythiourea as a catalyst for increasing molecular weight.
Japanese Kokai No. 55-56,129 published Apr. 24, 1980 discloses a process in which ammonium chloride is used as the catalyst in a reaction involving heating phosphazene oligomer at 150.degree.-350.degree. C. in a closed system. For example, a solution of linear and cyclic phosphonitrilic chloride oligomers in dichlorobenzene containing a small amount of ammonium chloride catalyst was heated at 255.degree. C. for 10 hours in a sealed tube to form the polymer.
Japanese Kokai No. 55-25,475 published Jan. 23, 1980 describes formation of phosphazene polymers by reacting a phosphorus source (e.g., P+Cl.sub.2 ; PCl.sub.3 +Cl.sub.2 ; PCl.sub.5) with a nitrogen source (e.g., NH.sub.3 ; NH.sub.4 Cl) in any of three reaction systems:
(1) In a solvent that does not dissolve the phosphazene polymers, such as an aliphatic hydrocarbon or alicyclic hydrocarbon that is resistant to halogenation. PA0 (2) In an undiluted (concentrated) reaction system having a small quantity (250 mL or less per mole of P source reactant) of a solvent capable of dissolving the phosphazene polymers that is resistant to halogenation, such as a halogenated aromatic hydrocarbon. PA0 (3) In a phosphazene oligomer as the solvent.
Japanese Kokai No. 55-65,228 published May 16, 1980 describes a method for producing phosphazene polymers in which a mixture of linear phosphazene oligomer, which has been stabilized with phosphorus pentachloride, hydrogen chloride or a metal halide, and cyclic phosphazene oligomer, is heated at 150.degree. to 350.degree. C. in a closed system having a solvent or non-solvent in the presence of a Lewis base catalyst. Urea, thiourea, polyurea, and polythiourea are examples of Lewis base catalysts used.
Japanese Kokai No. 55-50,027 published Apr. 11, 1980 discloses performing thermal ring-opening polymerization of cyclic phosphazene oligomers in the presence of linear phosphazenes stabilized with a metal halide, notably the linear oligomers formed as by-products when synthesizing the cyclic oligomers with metals or metal salts as catalysts. Such linear oligomers are indicated to have a degree of polymerization in the range of 2 to 100.
Japanese Kokai No. 55-60,528 published May 7, 1980 discloses a process wherein phosphazene polymers are formed by heating phosphazene oligomer at 150.degree. to 350.degree. C. in a closed system in the presence of a Lewis base such as urea, thiourea, polyurea or polythiorea. The phosphazene oligomer is a mixture of linear phosphazene oligomers (5 to 95 weight percent; stabilized with phosphorus pentahalide or hydrogen halide) and cyclic phosphazene oligomer.
Japanese Kokai No. 55-43,174 published Mar. 26, 1980 describes a process for producing phosphazene polymers in which cyclic phosphazene oligomers are subjected to thermal ring-opening polymerization in the presence of linear phosphazenes which have been stabilized by phosphorus pentahalides or hydrogen halides.
U.S. Pat. No. 4,374,815 describes a process wherein linear phosphonitrilic chloride polymers are produced from linear phosphonitrilic chloride oligomers by means of a two-step process. In the first step a mixture of linear phosphonitrilic chloride oligomer having an average degree of polymerization of at least four is heated with an excess of ammonia or ammonium chloride. Cyclic phosphonitrilic chloride oligomer formed during the course of the first step is removed from the reaction mixture. After the removal of the cyclic oligomer the reaction mixture is subjected to the second step which involves heating the mixture in an inert liquid solvent (optionally in the presence of ammonia or ammonium chloride) whereby the molecular weight of the linear phosphonitrilic chloride polymer is increased.
In copending application Ser. No. 447,720 filed Dec. 7, 1982, one of us (HML) describes an improvement in the process of U.S. Pat. No. 4,374,815 whereby the reaction rate when using ammonium chloride as the nitrogen source material in the first stage reaction is increased substantially by employing ammonium chloride having reduced particle size. For example, conversion of linear oligomer to linear polymer using ammonium chloride having a relatively small mean particle size of about 86 microns may be completed at a temperature of about 160.degree. C. after about one hour. On the other hand, when ammonium chloride having a relatively large mean particle size of about 115 microns is used, the polymerization rate at the same reaction temperature is slower and the yield of desired product is lower. The mean particle size of the ammonium chloride is preferably within the range of about 1 micron to about 100 microns and most preferably within the range of about 1 micron to about 90 microns.
In copending application Ser. No. 487,804 filed Apr. 22, 1983, one of us (HML) describes a process whereby linear phosphonitrilic chloride polymers are produced from phosphonitrilic chloride oligomers by means of a two-stage process. In the first stage a mixture of phosphonitrilic chloride oligomers enriched in linear oligomer is heated with at least a stoichiometric amount (and preferably an excess) of ammonia or ammonium chloride while concurrently removing hydrogen chloride to produce an intermediate reaction product. In the second stage the molecular weight of the reaction product is increased by heating the product to a higher temperature than the average temperature used in the first stage. Cyclic phosphonitrilic chloride oligomer may be present in the initial oligomer. Preferably, cyclic phosphonitrilic chloride oligomer is recovered from the reaction mixture after the start of the second stage. Most preferably it is recovered after completion of the second stage. At least a portion of the recovered cyclic oligomer may be used in a subsequent first stage reaction.
In copending application Ser. No. 487,805 filed Apr. 22, 1983, one of us (HML) describes a process whereby linear phosphonitrilic chloride polymers are produced from phosphonitrilic chloride oligomers by means of an improved process wherein a mixture of phosphonitrilic chloride oligomers enriched in linear oligomer is heated with at least a stoichiometric amount (and preferably an excess) of ammonia or ammonium chloride while concurrently removing hydrogen chloride to produce a higher molecular weight polyphosphonitrilic chloride product. Cyclic phosphonitrilic chloride oligomer may be and preferably is present in the initial oligomer. Preferably, cyclic phosphonitrilic chloride oligomer is recovered from the reaction mixture during (but more preferably after) completion of the polymerization reaction. At least a portion of the recovered cyclic oligomer may be used in a subsequent polymerization reaction.
In copending application Ser. No. 489,414 filed Apr. 28, 1983, one of us (HML) describes a process whereby halophosphazene polymers such as linear phosphonitrilic chloride polymers are prepared in two stages by (i) heating a nitrogen source (e.g., NH.sub.3 ; NH.sub.4 Cl; etc.) with an excess of a phosphorus source (e.g., PCl.sub.5 ; PCl.sub.3 +Cl.sub.2 ; P+Cl.sub.2 ; etc.) so that hydrogen halide is evolved and halophosphazene is formed, and (ii) heating at least a portion of the halophosphazene with at least a stoichiometric amount (preferably an excess) of a nitrogen source (e.g., NH.sub.3 ; NH.sub.4 Cl; etc.) so that hydrogen halide is evolved and halophosphazene polymer of higher molecular weight is formed. Preferably, a third stage is utilized wherein the resultant halophosphazene polymer is then heated, most preferably in an inert solvent or diluent which enhances the fluidity of the reaction mixture, to increase the molecular weight of the polymer.